Company_W

Company W wants to start testing a new software program within their four regions, Northeast, Southeast, Central and West. Each person in sales is required to have the same sales amount of product, however during the last 3months, only half of the sales representatives we given the software programs to help them manage their contacts. Due to this the VP of sales at Widgecorp wants to know all possible theories that can be put into place. This includes any statistical analysis that will explain why some of the sales representatives are not meeting their intended goals for the month. In order for a decision to be made on this issue, statistical testing must be completed in order for a more accurate conclusion can be made. There are different techniques that can be used for Company W to identify the statistical analysis to this matter. I will use the non-parametric statistics and hypothesis testing, these along with the chi-square distribution testing of data as the choices for this discussion. I first need to describe what each of these terms mean. Hypothesis Testing This is a technique that applies by businesses consecutively to acquire conclusions regarding population developing information taken from an example or sampling of a test data. The information taken from an example is collected so that a conclusion can be made by the analysis to either accept or reject the hypothesis. We have already discussed in other meetings, the null and alternative hypothesis, these terms fall into this category. As you might remember the null hypothesis is what the analysis is testing, trying to make the test false. In the conclusion of the test the analysis will either accept or reject the null hypothesis statement. The result will be the alternative hypothesis is the findings are false and rejected by the analysis. Non-Parametric Statistics This is an assessment that applies information categorically, resulting in a nominal or ordinal. The nominal pertains to one principle, and only one principle. They exist on one level of address such as gender, ethnicity, religion, nationality, or marital status. Nominal means “to name,” it’s a category that requires 1 answer category. You either fit in one category or the other. (Repurposed and updated from MGMT600-1304A-04, Dr. Pratl) Ordinal data is also based on one principle or level of address, but have a rank order. Measures are perfect examples of ordinal data in that they range from one extreme to another. Data collected on a scale of agreement (from strongly agree to strongly disagree) are ordinal data. The categories within the response set are strongly agreed, agreed, no opinion, disagree, strongly disagree. An order is implied in these data where one comes after another. If you were to mix up the order, it would not make sense. (Harry, et al, 2010) (Dockus, 2013, Repurposed and updated from MGMT600-1304A-04, Dr. Pratl) The nominal variables are presented as qualitative, while the ordinal variables are presented as qualitative and given a rank. The non-parametric analysis will not convey reports regarding information presented by the analyst. We will use the ANOVA, which is an analysis of variation, which is a common method used. By using this analysis it will tell if there are any variations among the groups and if the means of them are the same. So by using this technique we can determine is the null hypothesis will have the same means, while the alternative hypothesis will verify if the means are different. The analyst might use a one way- or two-way method of an ANOVA analysis. Which is defined as the one-way has only one factor to test the equality of the research or the two-way will differentiate if there might be other factors involved giving you every possible outcome to the situation within the observation. Chi-Square Distribution When variables are random they generally produce two types of information which are categorical and numerical. Most numerical scales are interval data. However, what distinguishes interval data from the final level is that there is no absolute zero in this scale. (Kennet & Salini, 2012) (Dockus, 2013, Repurposed and updated from MGMT600-1304A-04, Dr. Pratl) By using this technique we can discover the distinctions between the information collected and check if they are independent. Any variable that is specific and does not have any numerical value are considered categorical, while the numerical are just that, numerical. Meaning that is asked the question, that requires a number, would be numerical. So, if the study was asked how many sales did he produce this week, the answer would be in the form of a number, however, if he was asked what does his clients do for a living, that would fall under the categorical variable, due to the answer being a description rather than a number. Using the Chi-Square Analysis By using the null hypothesis, we can ask questions comparing the sales for the month were met by the sales representatives by using the new software compared to the sales met by the sales representatives that did not use the new software. The reality is that there is no proof that the sales representative that used the software met their sales for the month compared to the sales representatives that did not meet their goals by not using the new software. This hypothesis cannot be proven as true, which would make the alternative hypothesis the accepted fact; the same amount of software was not sold by the sales representatives. In conclusion testing a hypothesis is done by analysts to develop true statements regarding a problem or situation that can be accurately classified. By collecting, interpreting and analyzing data requires the analysts to have a firm understanding of the hypothesis that is being tested and find the resolve. Also the analyst needs to use additional techniques and theories in order to present to the company the findings and determine the best course of actions to their problem or situation. References Harry, Mikel J. & et al. (2010) Practitioner's guide for statistics and lean six sigma for process improvements [Books24x7 version] Available from http://common.books24x7.com/toc.aspx'bookid=33720 Kenett, Ron S. & Salini, Silvia (2012) Modern analysis of customer surveys: with applications using r. [Books24x7 version] Available from http://common.books24x7.com/toc.aspx'bookid=46186